Hello Zeek, I am not an aluminum expert, however is there a possibility that the edges of the materials that have been friction stir welded together have oxides/compounds that have been intermixed into the seam and when you are welding in that area you are volatizing those items? This is just a hunch and a simplified one at that. I am using the logic that since friction stir is not done in an inert shielded environment and isn't subjected to oxide removal through the action of the arc that the welds could contain certain items that might be gassing off when you weld over them. Look forward to hearing others responses and any conclusions that you come up with. My $.02. Best regards, Allan